Note: This is a summary of where to find files and analyses. README's with actual code are listed and described after each number
--Phenotype file: /nfs/beluga0_home/DATA/NIAGADS/NG00020/NG00020_LOAD_phenotype_files/NG00020_LOAD_phenotype/phenotype/2009_08_24_phenotype_data.csv --APOE genotypes: /nfs/beluga0_home/DATA/NIAGADS/NG00020/NG00020_LOAD_phenotype_files/NG00020_LOAD_phenotype/phenotype/2009_12_17 APOE results.csv --Age and sample source: /nfs/beluga0_home/DATA/NIAGADS/NG00020/NG00020_LOAD_phenotype_files/NG00020_LOAD_phenotype/phenotype/2009_11_16\ Age\ and\ Source\ File.csv (note: source of Sampled.Age and AgeAtLastEval variable... the phenotype file also has an AgeAtLastEval variable, but it is less complete)
--PC Relate GRM (by Jai): /nfs/beluga0_home/ANALYSIS/NIAGADS/NG00020/pcs_grm_pheno_only/out/b_pcrelate.rds --PC AiR PCs (by Jai): /nfs/beluga0_home/ANALYSIS/NIAGADS/NG00020/pcs_grm_pheno_only/out/b_pcair.rds
--gds: /nfs/beluga0_home/ANALYSIS/NIAGADS/NG00020/LOAD_families_010413_genotyped.gds
-Combines phenotype file, APOE genotypes file, and Age and sample sources file
-creates variables: -E2 (count) -E3 (count) -E4 (count) -E2_carrier -E4_carrier -E4_het_hom -censored_age (AAO for cases, AgeAtLastEval for controls) -surv (survival time) -Dev_Res (deviance residuals based on hazzard regression of "surv"... for AAO analysis)
-Extracts PCs 1-8 from b_pcair.rds and adds as variables (the 1st 4 PCs are use in analyses)
-generates annotated dataframe (/acct/hkings/NIAGADS/DATA/annot.rds)
-generates sample filters: -/acct/hkings/NIAGADS/DATA/keep_samples_noTwins.rds... samples excluded for: -no APOE genotype -1 individual per twin pair excluded based on PC Relate object (Note: file without twins excluded is called /acct/hkings/NIAGADS/DATA/keep_samples_e4_count) -No Case_Control status -/acct/hkings/NIAGADS/DATA/keep_samples_censored_analyses_noTwins.rds (This is the sample filter used in ALL analyses)... samples excluded for all above and also: -no AAO for cases -AAO after AgeDxDem for cases -AAO > 10 years before AgeDxDem for controls -no AgeAtLastEval for controls -AgeAtLastEval differs between phenotype and age_and_sample_source file
-Generates a variant sumamry file /acct/hkings/NIAGADS/DATA/variant_metrics.rds: -variant.id -missing.rate -ref.freq -maf
-Checks missingness rate for samples (none are < 0.05%)
-plots variant MAF, missingness, and sample missingness
-Generates a variant filter ~/NIAGADS/DATA/vars_qc_pre_HWE.rds: ~/NIAGADS/DATA/vars_qc_pre_HWE.rds (note: this is the exact same file as /acct/hkings/NIAGADS/DATA/vars_qc_keep.rds, which is used in assoc tests until HWE done) -maf < 0.05 -maf > 0 -missing.rate < 0.05 -HWE not calculated yet!!!
-calculated type I error... outputs saved to: /acct/hkings/NIAGADS/calculate_typeI_error/
files used in all analyses can be found here: /acct/hkings/NIAGADS/DATA/ and here: /nfs/beluga0_home/ANALYSIS/NIAGADS/NG00020/pcs_grm_pheno_only/out/
-annotated pehnotype file (with PCs): /acct/hkings/NIAGADS/DATA/annot.rds
-gds: /nfs/beluga0_home/ANALYSIS/NIAGADS/NG00020/LOAD_families_010413_genotyped.gds
-GRM: /nfs/beluga0_home/ANALYSIS/NIAGADS/NG00020/pcs_grm_pheno_only/out/b_pcr_mat.rds
-sample filter: ~/NIAGADS/DATA/keep_samples_censored_analyses_noTwins.rds
-variant filter (WILL BE UPDATED AFTER HWE DONE): ~/NIAGADS/DATA/vars_qc_keep.rds
config files, analyses, and follow-up annotation can be found in subfolders here: /acct/hkings/NIAGADS/analyses
-Analyses included in paper are (note: covars for all includes GRM): -AD (binary case/control); covars: PCs 1-4: /acct/hkings/NIAGADS/analyses/AD/AD_4PCs/assoc_readme.txt -AD (binary case/control); covars: PCs 1-4, E2 (count), E4 (count): /acct/hkings/NIAGADS/analyses/AD/AD_xE4E2_4PCs/assoc_readme.txt -AAO (DevRes continuous); covars: PCs 1-4: /acct/hkings/NIAGADS/analyses/AAO/AAO_4PCs/assoc_readme.txt -AAO (DevRes continuous); covars: PCs 1-4, E2 (count), E4 (count): /acct/hkings/NIAGADS/analyses/AAO/AAO_xE4E2_4PCs/assoc_readme.txt -E2_carrier; covars: PCs 1-4: /acct/hkings/NIAGADS/analyses/APOE/E2_carrier/E2_carrier_4PCs/assoc_readme.txt -E2_carrier; covars: none: /acct/hkings/NIAGADS/analyses/APOE/E2_carrier/E2_carrier_noPCs/assoc_readme.txt -E4_carrier; covars: PCs 1-4: /acct/hkings/NIAGADS/analyses/APOE/E4_carrier/E4_carrier_4PCs/assoc_readme.txt -E4_carrier; covars: none: /acct/hkings/NIAGADS/analyses/APOE/E4_carrier/E4_carrier_noPCs/assoc_readme.txt -E4_het_hom (NOT DONE); covars: PCs 1-4: /acct/hkings/NIAGADS/analyses/APOE/E4_het_hom/E4_het_hom_4PCs/assoc_readme.txt -E4_het_hom (NOT DONE); covars: none: /acct/hkings/NIAGADS/analyses/APOE/E4_het_hom/E4_het_hom_noPCs/assoc_readme.txt -I also referenced this in the paper: -sex; autosome-only; covars: PCs 1-4: /acct/hkings/NIAGADS/analyses/sex/sex_4PCs/assoc_readme.txt -sex; autosome-only; covars: none: /acct/hkings/NIAGADS/analyses/sex/sex_noPCs/assoc_readme.txt (note: analyses not listed above have not been kept up to date and may not utilize the same input files)
4. Analysis of GWAS hits and comparisons across analyses and to other papers: see assoc_readme.txt in each analysis subfolder
-Identifies lead SNPs below a P-value threshold (5x10^-5) for each GWAS -ex. output: /acct/hkings/NIAGADS/analyses/APOE/E2_carrier/E2_carrier_noPCs/SNP_peaks.txt
-Identifies and plots ranges surrounding each lead SNP (note: ranges are an estimate that should be checked visually) -ex. output: /acct/hkings/NIAGADS/analyses/APOE/E2_carrier/E2_carrier_noPCs/E2_carrier_noPCs_peak_ranges.txt
-Looks for overlap between E2/E4 (no PCs) GWAS hits (from peak_ranges files) and hits from other papers: -Andrews meta analysis: /acct/hkings/NIAGADS/DATA/andrews36_ranges.txt -see: /acct/hkings/NIAGADS/DATA/notes_on_Andrews.txt -ex. output: /acct/hkings/NIAGADS/analyses/APOE/E2_carrier/E2_carrier_noPCs/E2_carrier_noPCs_Andrews_intersect_ranges.txt -Grinde long-range LD regions: /acct/hkings/NIAGADS/DATA/Grinde36.txt -see: /acct/hkings/NIAGADS/DATA/notes_on_Grinde.txt ex. output: /acct/hkings/NIAGADS/analyses/APOE/E2_carrier/E2_carrier_noPCs/E2_carrier_noPCs_Grinde_intersect_ranges.txt
-Cross-analysis checks: /acct/hkings/NIAGADS/analyses/cross_analyses_tests_readme.txt -correlation between AD GWAS with and without APOE genotype control & correaltion between AAO GWAS with and without APOE genotype control -Look up lead SNPs from E4_carrier GWAS in AD and AAO (with and without APOE genotype control) GWAS -Look up lead SNPs from E2_carrier GWAS in AD and AAO (with and without APOE genotype control) GWAS
5. Data summary (generate table 1 / other data checks): /acct/hkings/NIAGADS/DATA/README_additional_data_summary_for_paper.txt
Summarizes results from prior GWAS of AD (that also control for APOE E2 and/or E4 in some way)
Tests correlation between these studies and plots Miami plots